From 78738d1616c15349c7c01dc3c62ace3d88af68fe Mon Sep 17 00:00:00 2001 From: harish Date: Thu, 10 Jun 2021 22:31:59 +0530 Subject: [PATCH 1/9] Example code for Spectrogram in documentation --- torchaudio/transforms.py | 14 ++++++++++++++ 1 file changed, 14 insertions(+) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index a04592f2c4..de564ab542 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -60,6 +60,20 @@ class Spectrogram(torch.nn.Module): dimension for real and imaginary parts. (see ``torch.view_as_real``). When ``power`` is provided, the value must be False, as the resulting Tensor represents real-valued power. + + Example:- + + >>> import torchaudio + >>> filename = "example.wav" + >>> waveform, sample_rate = torchaudio.load(filename) + >>> print("Shape of waveform: {}".format(waveform.size())) + Shape of waveform: torch.Size([2, 268237]) + >>> print("Sample rate of waveform: {}".format(sample_rate)) + Sample rate of waveform: 8000 + >>> specgram = torchaudio.transforms.Spectrogram()(waveform) + >>> print("Shape of spectrogram: {}".format(specgram.size())) + Shape of spectrogram: torch.Size([2, 201, 1342]) + """ __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized'] From e077d4baef7c29197a8be5000f217be82f3e8e25 Mon Sep 17 00:00:00 2001 From: harish Date: Fri, 11 Jun 2021 15:07:49 +0530 Subject: [PATCH 2/9] Spectrogram example resolved --- torchaudio/transforms.py | 17 +++++------------ 1 file changed, 5 insertions(+), 12 deletions(-) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index de564ab542..279c4bb870 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -61,18 +61,11 @@ class Spectrogram(torch.nn.Module): When ``power`` is provided, the value must be False, as the resulting Tensor represents real-valued power. - Example:- - - >>> import torchaudio - >>> filename = "example.wav" - >>> waveform, sample_rate = torchaudio.load(filename) - >>> print("Shape of waveform: {}".format(waveform.size())) - Shape of waveform: torch.Size([2, 268237]) - >>> print("Sample rate of waveform: {}".format(sample_rate)) - Sample rate of waveform: 8000 - >>> specgram = torchaudio.transforms.Spectrogram()(waveform) - >>> print("Shape of spectrogram: {}".format(specgram.size())) - Shape of spectrogram: torch.Size([2, 201, 1342]) + Example + >>> specgram = torch.randn(1, 40, 1000) + >>> specgram = torchaudio.transforms.Spectrogram()(specgram) + >>> specgram.shape + torch.Size([1, 40, 201, 6]) """ __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized'] From 86a7f815bcb10b02072c6f19d2eec04ac6349bc8 Mon Sep 17 00:00:00 2001 From: harish Date: Fri, 18 Jun 2021 09:07:37 +0530 Subject: [PATCH 3/9] Example Code for Spectrogram API --- torchaudio/transforms.py | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index de1a948c84..cec269e11b 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -69,10 +69,10 @@ class Spectrogram(torch.nn.Module): power spectrogram, which is a real-valued tensor. Example - >>> specgram = torch.randn(1, 40, 1000) - >>> specgram = torchaudio.transforms.Spectrogram()(specgram) - >>> specgram.shape - torch.Size([1, 40, 201, 6]) + + >>> waveform = torch.randn(1, 40, 1000) + >>> trans = torchaudio.transforms.Spectrogram(n_fft=800)(waveform) + >>> trans.shape """ __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized'] From d84f28775594a4949f51d4df3473e84c02e31ce2 Mon Sep 17 00:00:00 2001 From: harish Date: Fri, 18 Jun 2021 09:22:39 +0530 Subject: [PATCH 4/9] modified Spectrogram --- torchaudio/transforms.py | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index cec269e11b..8d16490736 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -70,9 +70,8 @@ class Spectrogram(torch.nn.Module): Example - >>> waveform = torch.randn(1, 40, 1000) + >>> waveform, sample_rate = torchaudio.load('test.wav', normalization=True) >>> trans = torchaudio.transforms.Spectrogram(n_fft=800)(waveform) - >>> trans.shape """ __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized'] From f3872ae6af06f021779bdb2af0820ad616c6260b Mon Sep 17 00:00:00 2001 From: harish Date: Fri, 18 Jun 2021 09:34:44 +0530 Subject: [PATCH 5/9] modified for space --- torchaudio/transforms.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index 8d16490736..1e9ee182ae 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -71,7 +71,7 @@ class Spectrogram(torch.nn.Module): Example >>> waveform, sample_rate = torchaudio.load('test.wav', normalization=True) - >>> trans = torchaudio.transforms.Spectrogram(n_fft=800)(waveform) + >>> trans = torchaudio.transforms.Spectrogram(n_fft = 800)(waveform) """ __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized'] From 8c49ed4bd2fdad668ed2f9315fe7ae7a3e701606 Mon Sep 17 00:00:00 2001 From: harish Date: Sat, 3 Jul 2021 11:13:57 +0530 Subject: [PATCH 6/9] modified Spectrogram --- torchaudio/transforms.py | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index 1e9ee182ae..5631b34964 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -68,10 +68,11 @@ class Spectrogram(torch.nn.Module): cases where ``power`` is a number as in those cases, the returned tensor is power spectrogram, which is a real-valued tensor. - Example + Example + + >>> waveform, sample_rate = torchaudio.load('test.wav', normalization=True) + >>> transformed_spectrogram = torchaudio.transforms.Spectrogram(n_fft = 800)(waveform) - >>> waveform, sample_rate = torchaudio.load('test.wav', normalization=True) - >>> trans = torchaudio.transforms.Spectrogram(n_fft = 800)(waveform) """ __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized'] From fe4a05b7dcf699d0d31ac1358c5075614ab99248 Mon Sep 17 00:00:00 2001 From: harish Date: Wed, 4 Aug 2021 09:33:52 +0530 Subject: [PATCH 7/9] Spectrogram --- torchaudio/transforms.py | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index 5631b34964..991a71f183 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -71,7 +71,8 @@ class Spectrogram(torch.nn.Module): Example >>> waveform, sample_rate = torchaudio.load('test.wav', normalization=True) - >>> transformed_spectrogram = torchaudio.transforms.Spectrogram(n_fft = 800)(waveform) + >>> transform = transforms.Spectrogram(n_fft = 800)(waveform) + >>> waveform = transform(waveform) """ __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized'] From ebcd94ad02d4da5429b8237bcc3700786fbe522b Mon Sep 17 00:00:00 2001 From: harish Date: Wed, 4 Aug 2021 09:37:21 +0530 Subject: [PATCH 8/9] spectrogram --- torchaudio/transforms.py | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index 991a71f183..53b27b9c95 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -71,8 +71,8 @@ class Spectrogram(torch.nn.Module): Example >>> waveform, sample_rate = torchaudio.load('test.wav', normalization=True) - >>> transform = transforms.Spectrogram(n_fft = 800)(waveform) - >>> waveform = transform(waveform) + >>> transform = torchaudio.transforms.Spectrogram(n_fft = 800) + >>> spectrogram = transform(waveform) """ __constants__ = ['n_fft', 'win_length', 'hop_length', 'pad', 'power', 'normalized'] From 44df364d97f2c66f27001baf722bffb38ab58cdd Mon Sep 17 00:00:00 2001 From: harish Date: Wed, 4 Aug 2021 10:50:05 +0530 Subject: [PATCH 9/9] Spectrogram --- torchaudio/transforms.py | 5 ++--- 1 file changed, 2 insertions(+), 3 deletions(-) diff --git a/torchaudio/transforms.py b/torchaudio/transforms.py index 53b27b9c95..c80a250b6e 100644 --- a/torchaudio/transforms.py +++ b/torchaudio/transforms.py @@ -69,9 +69,8 @@ class Spectrogram(torch.nn.Module): power spectrogram, which is a real-valued tensor. Example - - >>> waveform, sample_rate = torchaudio.load('test.wav', normalization=True) - >>> transform = torchaudio.transforms.Spectrogram(n_fft = 800) + >>> waveform, sample_rate = torchaudio.load('test.wav', normalize=True) + >>> transform = torchaudio.transforms.Spectrogram(n_fft=800) >>> spectrogram = transform(waveform) """